AI AND THE
FUTURE OF WORK

 

At the Institute, we are committed to promoting political and economic structures that leverage the power of AI to enhance innovation, creativity, interconnectivity and human flourishing. Technology can be used to empower or to exploit, and artificial intelligence (AI) is no exception. Who gets to benefit and who is exploited by technology is determined by the environment that is absorbing it: its political and economic structures, its power relations. Without well-structured and intentional intervention, those who lack power, particularly those from marginalized groups, are limited in their direct benefit from technological advancement, instead bearing the negative brunt of technological changes. 

Our goal is to shift technology use in the workplace and beyond from exploitation to empowerment.The dominant approaches to AI policy— regulating technologies after they enter the market and asking technology companies to voluntarily commit to social goals – are not enough to materially change AI’s trajectory. At the Institute, we draw upon an analytical framework that recognizes the inseparability of politics, economics, and identity for a more proactive, human-centered approach to AI policy. Technology development is shapeable; it’s responsive to its political and economic context. Together, we can bend the arc of AI toward worker empowerment and shared prosperity. 

The Status Quo 

Employers across the country are using new AI technology to tightly control work, devaluing labor to lower their own costs. This often manifests through: 

  • Deskilling workers and promoting narrow divisions of labor: Instead of empowering employees to solve a variety of problems, companies use AI to sew together jobs with a narrow division of labor. An example of this is in call center work, where voice recognition technology can route you to a call center worker who is empowered to resolve only one specific issue.  Because employers don’t need to invest as much in training, they are incentivized to adapt to high turnover rates rather than improve working conditions and compensation to retain workers.
  • Irregular hours and precarious income: Companies use scheduling technologies in conjunction with employment arrangements like part-time and zero-hour contracting to give maximum flexibility to the employer. The lack of guaranteed full time hours means more precarity for workers. 
  • Surveillance and algorithmic management: Companies use a combination of surveillance technology and algorithmic management to manage vast numbers of people engaged in high turnover work. At Amazon, pickers’ locations, item scanning speeds, and “time off task” are tracked, and can be used to automatically not renew their employment. Impacted jobs are subject to invasive surveillance, which has documented negative health effects. 

It is no coincidence that Black and Latinx workers are overrepresented in the kinds of low wage work discussed here. Subaltern identity groups are historically the ones who bear the cost of technological change, and indeed, this pattern bears out in contemporary research. 

Solutions

Most current efforts to address these challenges promote (1) policy change that relies on companies  to voluntarily go against the grain, or (2) policy work that treats the disenfranchisement of workers as an inevitable economic force and therefore focuses exclusively on after-the-fact policies to soften the blow of automation (e.g. workforce training and universal basic income). However, private employment arrangements are notoriously difficult to regulate, and expecting employers to voluntarily commit to empower these same employees is foolish. To change how technologies are absorbed into the workplace, we must change the incentives, constraints, and power relations that make exploitation attractive to employers today. At the Institute, we are calling for a paradigm shift in the AI and technology space. As an institute engaged in research, policy, and public engagement, we are uniquely capable to support the field in making this transition.

  • Research: Linking labor policy, identity group stratification, employment law, and corporate governance to AI development. Research suggests that certain worker protections– sectoral bargaining, expanding workers’ right to bargain over data collection and technology, and expanding definitions of employment under major employment law statutes – would have the cumulative effect of steering the development of AI technologies toward ones that enhance worker capabilities rather than replacing them. Our work will bring together scholars and researchers to demonstrate the causal relationship between technology use and labor policies, culminating in a special journal issue on a political economy approach to AI. 
  • Policy and Action: Applying research insights and bringing isolated work on AI policy in conversation and solidarity with the labor movement, identifying a shared way forward. The Institute hosts the National Jobs for All Network within its hub, and University Professor Darrick Hamilton also serves as the AFL-CIO’s Chief Economist. As such, we are ideally situated to translate our research findings into action. Additionally, due to our deep, existing relationships with technologists, policy makers, scholars, and the labor movement, we are well-positioned to bring tech and labor activists together in a series of convenings, resulting in actionable next steps. 
  • Public Engagement: Changing the narrative to proactive policy and hope. The current conversation on AI is pessimistic, and interventions are focused on preventing our worst fears about AI. Through a series of public events, speaking engagements, and op-eds, our work will bring a more hopeful, empowered focus on concrete policies that will steer the development of AI toward just, inclusive, and capability-enhancing ends. . 

We hope you’ll join us! If you’d like to learn more about how to support or get involved in our AI work, please reach out to Elaine Chang to learn more ([email protected]).